When cancer is identified at the earliest stages, the probability of cure is very high and therefore diagnostic screening tests that can detect these early stages are crucial. Efforts toward the development of early detection assays for cancers have traditionally depended on single biomarker molecules. Current technologies have been disappointing and have not resulted in diagnostic tests suitable for clinical practice. The core technology of this project rests on research from the PI's lab in which he has developed a high throughput method to identify large numbers of epitopes that can be used to identify the presence of breast cancer by detecting the presence of auto-antibodies to tumor proteins in the serum of the test subject. These biomarkers are cloned without a preconceived notion of their function and can be used as diagnostic and prognostic biomarkers. The essential features of the approach are acknowledging the heterogeneous nature of any specific kind of cancer, departing from the reliance on any single marker for disease detection, and using specialized bioinformatics techniques to interpret the results. The concept employs pattern recognition of multiple markers as a diagnostic rather than any single marker. This study is possible due to a remarkable specimen bank consisting of primary breast tumor tissues and sera collected frozen at the time of surgery from 1306 patients from Detroit metropolitan hospitals from 1975 to 1983. In addition, follow-up sera were collected from participating patients through 1992. Interview follow-ups were continued until 1996 and data forms from patient follow-ups are available. Over 11,000 sera have been maintained at -70[unreadable] C at the KCI facility and are still currently available for all primary samples. In the discovery phase, the serum antibodies have been detected by screening of large numbers of potential epitope targets on protein microarrays. By a unique combination of techniques to enhance detection (microarrays of bacteriophage-bearing displayed tumor antigens), this study proposes to investigate a novel serum assay to detect and predict outcomes for breast cancer. The principle is that we can clone epitopes reacting with IgG in patients sera and use them to detect antibodies in sera to discriminate cancer and healthy subjects and detect disease prior to standard diagnosis. Such an approach should provide an early detection assay for asymptomatic women. [unreadable] [unreadable]